ERADICATION OF BLACK RATS FROM FARALLÓN …iws.org/CISProceedings/7th_CIS_Proceedings/Samaniego-Herrera_et_al.pdferadication of black rats from farallÓn de san ignacio and san pedro

Pages 337–347 in Damiani, C.C. and D.K. Garcelon (eds.). 2009. Proceedings of 337 the 7th California Islands Symposium. Institute for Wildlife Studies, Arcata, CA.

ERADICATION OF BLACK RATS FROM FARALLÓN DE SAN IGNACIO AND SAN PEDRO MÁRTIR ISLANDS, GULF OF CALIFORNIA, MEXICO

ARACELI SAMANIEGO-HERRERA,1 ALFONSO AGUIRRE-MUÑOZ,1 GREGG R. HOWALD,2 MARIA FÉLIX-LIZÁRRAGA,1 JORGE VALDEZ-VILLAVICENCIO,1

RICARDO GONZÁLEZ-GÓMEZ,1 FEDERICO MÉNDEZ-SÁNCHEZ,1 FLOR TORRES-GARCÍA, MARLENNE RODRÍGUEZ-MALAGÓN,1 AND BERNIE R. TERSHY3

1Grupo de Ecología y Conservación de Islas, A.C., Av. López Mateos No. 1590-3, Fracc. Playa Ensenada, Ensenada, B.C. 22880, México; [email protected]

2Island Conservation Canada, 680-220 Cambie Street, Vancouver, BC, V6B 2M9, Canada3University of California Santa Cruz, Center for Ocean Health, 100 Shaffer Road, Santa Cruz, CA 95060

Abstract—In the Gulf of California, Mexico, several islands have been severely impacted by introducedrats (Rattus spp.). A rat eradication project for Farallón de San Ignacio and San Pedro Mártir islands, bothglobally important seabird colonies, unfolded from planning and baseline studies in 2005 toimplementation in 2007. Rats were eradicated via aerial helicopter broadcast of bait, the only method usedto date to eradicate rats from large and topographically complex islands. A special aerial bucket developedin New Zealand (Helicopters Otago) was used for the helicopter dispersion of CI-25, a special formulationof Brodifacoum in the form of pellets, provided by Bell Labs. Eradication occurred in September 2007 onFarallón de San Ignacio and in October–November 2007 on San Pedro Mártir. Initial post-eradicationmonitoring resulted in no sign of rats. There were no losses of native fauna at a population level. Post-eradication monitoring will continue throughout 2008–2009 to confirm both rat absence and ecosystemrecovery. Interinstitutional cooperation was essential in achieving complex permitting, logistics, andfinancial challenges. This project represents the first project of aerial broadcast rat eradication in LatinAmerica.

INTRODUCTION

Introduced vertebrates are known to be a majorcause of extinctions and other dramatic changes onislands (Mack et al. 2000; Blackburn et al. 2004;Reaser et al. 2007). In particular, the impact of rats(Rattus spp.) on native island plants, invertebrates,and vertebrates has been well documented(Whitaker 1973; Campbell 1991; Cree et al. 1992;Navarrete and Castilla 1993; Lee and Yoo 2002;Towns et al. 2006). Rats are present on about 85% ofthe world’s island chains (Atkinson 1985) and havecaused 40%–60% of all bird and reptile extinctionssince 1600 (Groombridge 1992).

Two islands that have been impacted by ratintroductions are Farallón de San Ignacio (FSI) andSan Pedro Mártir (SPM) islands in the Gulf ofCalifornia, México (Fig. 1). Both islands supportseabird colonies of national and internationalimportance (Tershy and Breese 1997; González

Figure 1. Locations of Farallón de San Ignacio and San PedroMártir islands, México.

338 SAMANIEGO-HERRERA ET AL.

Bernal et al. 2002), are part of a World Heritage Site(UNESCO), and are legally protected by Mexicanlaw as Natural Protected Areas. Black rats (Rattusrattus) are thought to have been accidentallyintroduced to both islands during guano miningactivity in the early 1900s (Tershy et al. 1997).Precise information about local impacts ofintroduced rats on these islands is scarce (Velardeand Anderson 1994). However, rats are known tohave impacted island populations through predationon seabird eggs and chicks on both islands, andfruits and flowers of the cardón (giant cactus;Pachycereus pringlei) on SPM (Tershy et al. 1997;A. Samaniego, personal observation). Additionally,the absence of nocturnal seabirds despite suitablehabitat on both islands, limited distribution offishing bats (Myotis vivesi) on SPM, and extremelylow abundance of the gecko (Phyllodactylushomolepidurus) on FSI, are all considered probableconsequences of rat predation (Tershy et al. 1997;González Bernal et al. 2002; Frick and Heady 2007;Peralta-Garcia et al. 2007).

To promote the natural restoration of FSI andSPM islands, a rat eradication plan was developed in2005. The project was headed by the Mexicanprivate organizat ion Grupo de Ecología yConservación de Islas (GECI), in collaboration withMexican government agencies SEMAR, SEGOB,CONANP, and SEMARNAT. GECI´s sisterorganizat ions in the United States (IslandConservation) and Canada (IC Canada), as well asPrescott College in Bahía Kino, were the maincollaborators in fundraising, planning, and localsupport, respectively. Prior to 2006, four rateradications had been attempted on Mexican islandsby putting out stations containing poison baits(Tershy et al. 2002). Three of these eradications(Rasa, San Jorge, and San Roque islands) weresuccessful (Tershy et al. 2002), and one of them(Isabel Island) failed (Rodríguez et al. 2006). Allthree islands with successful eradications were lessthan 50 ha in area, whereas Isabel Island was 82 ha,suggesting that the bait station technique may noteffectively eradicate rats on larger islands where notall rats can access the bait stations. Because FSI andSPM islands are topographically complex and SPMis large (267 ha), we opted to use a helicopter todisperse rodenticide broadly across each island.Although used effectively elsewhere (Howald et al.2007), this was the first time that this aerial

procedure was applied in Latin America. The rateradication project therefore included a 2-year pre-eradication phase in which we acquired baselinepopulation data on rats, and conducted studies to testwhether rodenticide pellets were likely to beconsumed by native vertebrates. Here we presentthe results of these preliminary studies, describe theaerial application of rodenticide to each island, andreport initial results of post-application monitoring.

METHODS

Site DescriptionFarallón de San Ignacio (25° 26’ 11.5” N, 109°

22’ 45.5” W) is a small island (17 ha) located 27 kmoff the coast of Sinaloa state, México (Fig. 1). It is atall rock with vertical walls and a flat top at 137m.a.s.l. (Fig. 2a). Vegetation is completely non-existent. The island is an important seabird nestingsite. Blue-footed booby (Sula nebouxii), brownbooby (S. leucogaster brewsteri), red-billedtropicbird (Phaeton aethereus), and Heermann´s

Figure 2. Permanent trapping transects on (a) Farallón de SanIgnacio and (b) San Pedro Mártir islands.

ERADICATION OF BLACK RATS 339

gull (Larus heermanni) are the most abundantspecies (González-Bernal et al. 2002). Besidesseabirds, the other native vertebrates are threespecies of lizards (Aspidoscelis tigris, Urosaurusornatus, and Phyllodactylus homolepidurus;González Bernal et al. 2001a; González Bernal et al.2001b; Peralta-García et al. 2007) and the Californiasea lion (Zalophus californianus). No nativeterrestrial mammals occur on the island (CONANP2000; Case et al. 2002). The only introducedmammal was the black rat R. rattus (González-Bernal et al. 2002).

Located 60 km off the coast of Sonora state,México, San Pedro Mártir (28° 23’ 0.0” N, 112° 18’30.0” W) is the most isolated island in the Gulf ofCalifornia (Fig. 1). The island is 267 ha with amaximum altitude of 305 m.a.s.l., and is dominatedby mountainous cliffs (Fig. 2b). It hosts 27 speciesof plants (cardón forest being the predominantvegetation), 53 of terrestrial birds, 36 of seabirds (8nesting), 4 of reptiles, and 1 of pinnipeds (Tershyand Breese 1997; Grismer 2002; CONANP 2007).SPM supports some of the biggest populations in theworld of blue-footed boobies (S. nebouxii), brownboobies (S. leucogaster brewsteri), and red-billedtropicbirds (P. aethereus). No native terrestrialmammals occur on the island (Case et al. 2002;CONANP 2007). The only introduced mammal wasthe black rat R. rattus (CONANP 2007).

Pre-Eradication Rat MonitoringTo collect baseline data on rat populations prior

to application of rodenticide, we established threepermanent transects on each of the two islands inMay 2005. On SPM the three transects representedthe three main habitat types: coast, canyon andinland top (Fig. 2b). Because it was logisticallyunfeasible to trap along the coast on FSI, weestablished two transects on the inland top and onetransect along a steep canyon (Fig. 2a). Eachtransect included 15 trapping points at intervals of20 m. At each point we set 3 devices separated by 2m: one Tomahawk trap (Tomahawk Live Trap Co,Tomahawk, WI), one Sherman trap (H.B. Shermantrap, Tallahassee, FL), and one indicator block(peanut-flavored wax chew block). Especially forislands without native rodents, indicator blocks are avery useful, inexpensive and practical tool fordetecting rodent presence. Traps and blocks wereset and checked for three consecutive days at each

trapping session. All traps were opened beforesunset, baited with a mix of oats and peanut butter,and checked in the morning. Indicator blocks wereset the first afternoon, checked every morning andreplaced if they showed any mark. Data recordedincluded island, date, transect, trapping point, traptype, species captured, and marks on the block(none, rodent bites, and other marks).

Systematic trapping combined with indicatorblocks was conducted every three months on bothislands from fall 2005 to summer 2007, except fall2006 for FSI and winter 2006 for both islands. Thisresulted in one trapping period in fall (2005), one inwinter (2005), two in spring (2006, 2007), and twoin summer (2006, 2007) on FSI; and two trappingperiods in fall (2005, 2006), one in winter (2005),two in spring (2006, 2007), and two in summer(2006, 2007) on SPM. On FSI total capture effortrepresented 1319 trap-nights and 475 block-nights.On SPM total capture effort represented 1513 trap-nights and 435 block-nights. Percent capturesuccess for each trapping period was calculatedaccording to the number of individuals captured pertrap-nights (Tomahawk and Sherman trapscombined) given habitat type and trapping date. Allcaptured individuals (100% R . rattus) werehumane ly sac r i f i ced wi th t he anes the t i cPentobarbital sodium (Aranda Labs, Querétaro,Querétaro, México).

Choice of Rodenticide and Potential Impacts onNative Fauna

Bait used was the rodenticide CI-25, developedby Bell Labs (Madison, WI) especially forecological restoration projects, and provensuccessful on Anacapa Island in the Channel Islandsin 2001 and 2002 (Howald et al. 2005). CI-25 aregreen, unwaxed, compressed grain, 2 gm pelletscontaining 25 ppm brodifacoum, which is a secondgeneration anticoagulant. Prior to rat eradicationand in order to assess if CI-25 pellets were attractiveto island fauna, we conducted palatabili tyexperiments of placebo bait on reptiles, seabirds,and fishes (Table 1) on April 3–7, 2006 on FSI andon April 11–16, 2006 on SPM. Placebo bait was alsomanufactured by Bell Labs and had the samecharacteristic of the toxic bait except for therodenticide brodifacoum.

We tested all reptile species whose diet couldinclude grain-type components, and excluded


Table 1 . Species, methods, and reaction of vertebrates exposed to placebo bait pellets of CI-25 (brodifacoum) on Farallón de SanIgnacio (FSI) and San Pedro Mártir (SPM) islands.

Species Family Island Method of exposure Reaction

Fishes

Prionurus punctatus Acanthuridae FSI Natural conditions No consumption

Balistes polylepis Balistidae SPM Natural conditions No consumption

Ophioblennius steindachneri

Blenniidae FSI, SPM

Natural conditions Inspection, no consumption

Cirrhitus rivulatus Cirrhitidae FSI Natural conditions No consumption

Diodon holocanthus Diodontidae FSI Natural conditions No consumption

Haemulon sexfasciatum Haemulidae FSI, SPM

Natural conditions No consumption

Girella simplicidens Kiphosidae SPM Natural conditions No consumption

Bodianus diplotaenia Labridae SPM Natural conditions No consumption

Semicossyphus pulcher Labridae SPM Natural conditions No consumption

Thalassoma lucasanum Labridae FSI, SPM


Lutjanus argentiventris Lutjanidae SPM Natural conditions No consumption

Lutjanus novemfasciatus Lutjanidae SPM Natural conditions No consumption

Holacanthus passer Pomacanthidae FSI, SPM


Microspathodon dorsalis Pomacanthidae SPM Natural conditions No consumption

Stegastes acapulcoensis Pomacentridae FSI Natural conditions No consumption

Stegastes flavilatus Pomacentridae FSI Natural conditions No consumption

Abudefduf troschelii Pomacentridae FSI, SPM


Chromis atrilobata Pomacentridae FSI, SPM


Cephalopholis panamensis

Serranidae SPM Natural conditions No consumption

Epinephelus labriformis Serranidae FSI Natural conditions No consumption

Mycteroperca rosacea Serranidae SPM Natural conditions No consumption

Paranthias colonus Serranidae FSI, SPM


Axoclinus carminalis Tripterygiidae SPM Natural conditions No consumption

Reptiles

Urosaurus ornatus Phrynosomatidae FSI Natural conditions Inspection, no consumption

Uta palmeri Phrynosomatidae SPM Natural conditions and captivity Inspection, no consumption

Aspidoscelis tigris Teiidae FSI Natural conditions and captivity Inspection, no consumption

Aspidoscelis martyris Teiidae SPM Natural conditions Inspection, no consumption


species with strictly carnivorous diets (e.g.,Phyllodactylus homolepidurus, Lampropeltisgetula nigritus, and Crotalus atrox). Twenty adultlizards (10 Aspidoscelis tigris on FSI and 10 Utapalmeri on SPM) were individually kept in captivityfor 5 days in wired cages (90 x 45 x 45 cm) on theirrespective island. Each individual was offered a 2.0gm pellet, which was subsequently weighed andchecked for bite signs every 24 hrs. Along with theexperiment, we made in situ observations on adultand juvenile individuals of four species of lizards(A. tigris and Urosaurus ornatus on FSI; Utapalmeri and A. martyris on SPM). Between 10:00and 11:00 a.m., we spread out 10 pieces of placebobait inside plots of approximately 10 m2 in differentmicrohabitats (shoreline, canyon, inland top). Ineach trial (n = 15 per island), we observed thereaction of each species for 15–30 minutes.Reactions to pellets were categorized as a) noconsumption (pellets ignored), b) inspection but notconsumption, or c) consumption.

For birds we conducted in situ observations inorder to include the most species possible. Becauseterrestrial birds are present at very low abundance,they were never present during our observations.Despite the fact that most of the nesting seabirds arepiscivorous, we were interested in the reaction ofground nesting seabirds to bait pellets. Observationswere conducted during morning hours on areas withabundant individuals (> 50) for five consecutivedays on each island. In each trial (n = 15 per island),we placed 10 pieces of placebo inside plots ofapproximately 50 m2. Reaction of each species wasrecorded during 30–60 minutes. Reactions to pelletswere categorized as a) no consumption (pelletsignored), b) inspection but not consumption, or c)consumption.

Since the probability of spreading bait beyond10 m offshore of the islands was minimal, wefocused our f i sh obse rva t ions on wa te r simmediately adjacent to the islands. Sevenunderwater observation sessions were conducted(four on FSI and three on SPM). Each session tookplace in a different day, during the afternoon, andlasted for an hour. At each session one person withfree diving equipment remained 3–5 m from shoreand made observations along the water column up to15 m deep. In each event, 10 pieces of placebo werethrown to the ocean. Species present and reaction topellets were recorded. Reactions were categorizedas a) no consumption (pellets ignored), b) inspectionbut not consumption, or c) consumption.

Bait ApplicationBecause of the size, steepness, and ruggedness

of FSI and SPM islands, the most feasible option forachieving eradication was to disperse bait pelletsusing an aerial drop technique, developed in NewZealand (Towns and Broome 2003) and usedpreviously in several countries (Howald et al. 2007).The bait was broadcast from a helicopter using astainless steel spreader bucket built in New Zealand(Helicopters Otago, Mosgiel, New Zealand). Thehelicopter (Bell 206 from Aspen Helicopters,Oxnard, CA) was equipped with a differential GPSto obtain geographic data with high accuracy. AGIS was built to confirm that distribution andapplication rate within each topographic categoryarea were correct according to planning. On bothislands bait was applied to 100% of the islandsurface. Cliff and canyon areas were treated twiceduring each drop to allow for the increased planararea that had to be covered. When applying bait tothe cliffs, a lateral deflector was adapted on thebucket to narrow the angle of bait dispersion and

Species Family Island Method of exposure Reaction

Birds

Larus heermanni Laridae FSI, SPM


Sula leucogaster brewsteri Sulidae FSI, SPM


Sula nebouxii Sulidae FSI, SPM


Table 1 (continued). Species, methods, and reaction of vertebrates exposed to placebo bait pellets of CI-25 (brodifacoum) onFarallón de San Ignacio (FSI) and San Pedro Mártir (SPM) islands.


minimize the amount of bait spread into the ocean.Because the intertidal zone represents prime habitatfor rats, a detailed hand broadcast was conducted bysmall boat (along shoreline) and by helicopter(above islets) the next day after each aerial drop toensure adequate bait application. To track the aerialwork and map rodenticide densities in the ground,we used a 60 cm per pixel resolution Quickbirdsatellite image (© 2006 Digital Globe Inc.) of SPM,and an aerial photo of FSI.

Climate and reproductive cycles of bothintroduced and native species are the main factors tobe taken into account for timing rat eradications(Howald et al. 2005). In this case, timing of baitapplication was determined by weather (dryseason), seabird and sea lion activity (non-reproductive season), and rat population activity(low breeding rate). The eradication operation onFSI was carried out on September 27, 2007, with thehelicopter operating out of the Mexican Navy basein Topolobampo, Sinaloa, 29 km from the island. Atotal of 567.5 kg of bait was broadcast by helicopterusing a single aerial bait drop and 90 kg of bait wasbroadcast by hand along shoreline and islets. Onaverage, bait was applied at a rate of 24.4 kg/ha (Fig.3a). The operation on SPM included two bait drops.The operation base for the first drop was theMexican Navy MV Sonora (PO152), an oceanicpatrol vessel from the Mexican Navy provided forthe project that anchored in front of the island. Thedrop occurred on October 31, 2007. The seconddrop was based on the island and occurred nine dayslater on November 9, 2007, with logistics supportfrom the MV Guadalupe, also from the MexicanNavy. Average bait density broadcast was 17.6 kg/ha (Fig. 3b). A total of 5,902.5 kg of bait wasbroadcast by helicopter and 440 kg by hand alongshoreline and islets.

Underwater monitoring was conducted 24hours after each aerial broadcast, in both theintertidal and sublittoral zones of the two islands.The dive focused primarily in assessing theinvertebrate and fish communities, as well as the sealion colonies. Underwater photographs were taken.

Confirmation MonitoringThe first eradication confirmation monitoring

on FSI was carried out seven weeks after the drop,on November 14–17, 2007. 105 Tomahawk trapsand 260 indicator blocks were set along permanent

transects and around areas with known high ratactivity, for three consecutive nights. Capture effortrepresented 315 trap-nights and 780 block-nights.All accessible areas were walked searching forcarcasses (of rats or native species) and signs of rats.

On SPM, it was not possible to return weeksafter the drop due to financial and logisticlimitations. Therefore, the initial confirmation wasbased on monitoring of radio-collared rats. Twelveadult females (average weight = 156.2 gm ± 30.3)and 13 adult males (average weight = 178.2 gm ±36.5) were captured in different habitat types andradio-collared 12 days before the first drop. All werereleased at the capture spot and monitored everynight . Based on te lemetry indicat ions, 24individuals were active the day of the first drop. Onecollar may have become damaged because thesignal was lost days before the drop. After the drop,individuals inactive for 2–3 days were assumed tobe dead and were subsequently recovered fromunderground burrows. Date of death was estimatedbased on carcass condit ion. All recoveredindividuals were dissected to confirm cause of

Figure 3. Bait density after (a) a single broadcast on Farallónde San Ignacio Island and (b) a double broadcast on San PedroMártir Island, in fall 2007.


death. Nine days after the first drop, we walked five500 m transects spread over island, looking forcarcasses (of rats or native species) and signs of rats.Thirty-two circular monitoring plots of 3 m radius(28.27 m2), chosen randomly around the islandbefore the drops, were searched in greater detail tomaximize the probability of finding small animals.

RESULTS

Rat Monitoring Prior to the application of rodenticide, capture

success varied between islands, seasons, and habitattypes (Fig. 4). On FSI, 238 individuals of R. rattuswere caught from a total capture effort of 1319 trap-nights (18.0% capture success). The highest meanseasonal capture rate was in summer ( 29.8%;Fig. 4a), and the highest mean capture rate amonghabitat types was in inland top ( 18.1%; Fig. 4a).38.1% of all indicator blocks were chewed by rats.On SPM, 252 individuals of R. rattus were caughtfrom a total capture effort of 1513 trap-nights(16.7% capture success). The highest seasonal meancapture rate was in summer ( 32.1%; Fig. 4b). Ofthe habitat types, the highest mean capture rate wasin canyon ( 25.1%; Fig. 4b). Rats chewed 21.1%of all indicator blocks.

On FSI, preliminary trapping followingapplication of bait yielded no captures of rats, andthere was no evidence of rat chewing on indicatorblocks. No rat carcasses were found duringsearching walks. On SPM, 19 of 24 radio-collaredrats were recovered dead, along with one droppedcollar. Of the remaining four collared rats, threewere located in inaccessible locations (vertical cliffsand under massive rocks); and one, including theactive collar, was devoured by a rattlesnake. Allrecovered rats died underground between 3 and 7days after the first drop. Dissection of 21 individuals(19 collared plus 2 non-collared found whiledigging for the first ones) confirmed poisoning asthe cause of death in 100% of the rats. Furthermore,stomach content in most individuals was 90% bait.During searching walks and inspections ofverification plots, we found two fresh rat carcasses.

Impacts on Native FaunaTable 1 shows a summary of palatability

experiments. None of the reptiles in captivity

consumed bait; all pellets weighed the same at theend of the experiment as at the beginning, and noneshowed bites. As for the in situ observations, theonly reaction recorded for all species and islandswas “inspection but not consumption.” Seabirdspecies present during the observations were S.nebouxii, S. leucogaster brewsteri, and Larusheermanni. The only reaction recorded for allspecies and islands was “not consumption.” On FSI13 fish species of 9 families (Table 1) were recordedduring the observations, whereas on SPM 17 speciesof 10 families (Table 1) were recorded. Of a total of23 fish species observed, 22 did not consumepellets. For the last species (Ophioblenniussteindachneri), one individual out of a dozenobserved ate one piece, but spit it out. The samespecies was observed on SPM, but was not observedto consume bait pellets.

Twenty-four hours following application ofrodenticide, underwater monitoring at both islands

=x

=x

=x

=x

Figure 4. Percentage of capture success of pre-eradication rattrapping per habitat type, on (a) Farallón de San Ignacio Islandand (b) San Pedro Mártir Island, from 2005–2007. Heights ofbars indicate means, and error bars indicate maximum andminimum values.


found no traces of pellets on either the sea floor or intide pools. Five species of invertebrates: red rockcrabs (Grapsus grapsus), barnacles (Balanus spp.),sea urchins (Echinometra vanbrunti), and sea stars(Pharia pyramidata and Phataria unifascialis); 13fish species of 9 families (Table 1), and sea lions (Z.californianus), were recorded in normal conditionsin terms of behavior and external aspect.

Seven weeks af ter the drop, advanceddegradation of leftover bait was evident. On theinland´s top of FSI, 11 small terrestrial birds (8Passer domesticus, 2 Columbina passerine, and 1Carpodacus mexicanus) were found dead, possiblyby rodenticide ingestion. Carcasses were too dry todissect and confirm cause of death. Blue-footedboobies, brown boobies, and red-billed tropicbirdshad laid the first eggs of the season and all wereintact, without rat chews. During searching walksand inspections of verification plots on SPM, wefound seven fresh bird carcasses: six yellow-footedgulls (Larus livens) and one common raven (Corvuscorax), which may have died from poisoning. Alllizards, snakes, and other birds observed during thesearching walks looked in normal condition in termsof behavior and external aspect.

DISCUSSION AND CONCLUSIONS

Two years of planning, testing, and monitoringwere necessary to obtain the basic information fordevelopment of a detailed strategy for eradicatingrats on FSI and SPM islands, and to comply withcomplex permitting requirements. Baseline dataacquired from pre-eradication monitoring facilitatesthe confirmation of both eradications, as well as theplanning of other eradication and conservationactions in the region. As expected, rats on SPM diedbetween days three to seven after the first bait drop.Initial results indicate that the rat eradications onboth FSI and SPM islands were successful.

We believe that application of the rodenticideCI-25 to FSI and SPM islands resulted in minimaladverse effects to island fauna. Results ofpalatability experiments with the abundant nativespecies suggest that bait pellets intended for rats arenot likely to be attractive to many native fish,reptiles, or seabirds. Of the 29 vertebrate speciesexposed to the placebo bait, 96.5% were notinterested in it, and the few that were attracted (three

reptiles, one fish) did not consume even one pellet.Nevertheless, because of their migratory habits,behavior, and potential susceptibility, birds needspecial attention and more studies. Overall, the lossof native fauna after the broadcast was insignificantat a population level. However, some mortalitieswere observed during monitoring that may haveresulted from rodenticide poisoning: 11 terrestrialbirds (3 species of passeriforms) on FSI as well as 7seabirds (6 yellow-footed gulls and 1 commonraven) were found dead. None of these species areabundant on the islands and were not present duringour palatability tests. Since these passeriformspecies are granivorous, and seagulls and ravens aregeneralists, primary poisoning may have been thecause. Seagulls and ravens are also scavengers andtherefore are exposed to secondary poisoning aswell. At the same time, since rats die undergroundthey are not expected to be available for scavengers,and insects are not susceptible to anticoagulants.Therefore, impacts via secondary poisoning areexpected to be minimal. However, a detailed andspecific plan for each case, containing mitigationactions for both marine and terrestrial ecosystems, isobligatory. For example, during the black rateradication in 2001–2002 on Anacapa Island,endemic mice were taken into captivity and raptorswere temporarily maintained at a prudent distancefrom the island (Howald et al. 2005).

Following application, no traces of pellets orsigns of negative environmental impact were foundat the intertidal and sublittoral zone of FSI and SPMislands, and it appears that the deflector wassuccessful in minimizing the number of bait pelletsbroadcast into the ocean. Consistent with otheraerial rat eradications (Howald et al. 2005; Hoareand Hare 2006), marine invertebrate and fishcommunities did not appear to be negativelyaffected by the low amount of bait that fell into thewater. No signs of poison-caused death were foundin marine species. Helicopter disturbance to faunawas inconsequential, as the operation lasted only afew days and did not occur during the reproductiveseason for seabirds or pinnipeds.

Conventionally, formal protocols on rodenteradications establish that success can be declaredafter two years of the eradication, since rodents atlow densities are difficult to detect (Howald et al.2007). Post-eradication monitoring on these islandswill continue seasonally, every three months. Once


the eradication is fully confirmed in fall 2009, thelong-term monitoring will focus in those habitatsand seasons where rats used to be more abundant.We expect post-eradication monitoring throughout2009 to confirm both rat absence and ecosystemrecovery. This project represents the first aerialbroadcast rat eradication in Latin America and thesecond in North America.

Interinstitutional cooperation has been key toachieving the complex permitting, logistic, andfinancial challenges. The Mexican Navy, wellabove and beyond its mandate, has been activelyengaged during the last years in island conservationprojects (Aguirre-Muñoz et al. 2008). In this case,the Navy supported the helicopter operation byproviding a modern vessel with helicopter platformand hangar as an in-kind donation. Also, the projectgot the backing from the Navy’s facilities and thepersonnel on the mainland. The Ministries of theIn te r io r (Gobernac ión) and Envi ronment(SEMARNAT) have moved from a traditionalregulatory and supervising role to a more proactiveone (Aguirre-Muñoz et al. 2008). The teams in thefield had the participation of technical personnelfrom the Protected Areas Commission (CONANP);for the CI-25 special bait importation, theCONANP’s head, Ernesto Enkerlin personallyassumed responsibility in front of customs andhealth authorities. This intense teamwork between aprivate organization and federal agencies is theresult of trust and confidence gained during severalyears, honored commitments, and shared successfulresults. Because of the logistical complexities ofworking on islands, and once Mexican islands arefederal territories, for future similar projects inMexico the same commitment and support ofMexican government agencies is a must.

Based on what we can define as a focusedcollaboration approach, the “winning strategy” foreradicating 43 introduced mammal populations on28 Mexican islands during the last decade integrates3 main elements: (1) a persistent and capable privateorganiza t ion leading the process wi thoutdistraction, free of bureaucratic limitations, andretaining its specialized personnel; 2) sufficient andt imely funding coming from nat ional andinternational sources; and 3) the proactiveengagement of federal authorities. Internationalcollaboration has proven also to be important, as

other countries such as New Zealand have vastexperience in this subject.

This particular eradication project will benefitmany desert island plants and invertebrates, sevenspecies of reptiles, more than nine species ofseabirds, and fishing bats, and will represent anenormous advance regarding rat eradications inMéxico. There are at least 20 more islands inMexico where ae r ia l p rocedures mus t beimplemented in order to eradicate introducedrodents. Keeping and consolidating the currentpace, the complete restoration of all 24 remainingMexican islands with 60 populations of introducedmammals now appears to be a viable strategic goalthat will be achievable by 2025.

ACKNOWLEDGMENTS

Main logistics would be never been solvedwithout the very generous collaboration of theSecretaría de Marina-Armada de México (MexicanNavy). The Secretaría de Gobernación and theComisión Nacional de Áreas Naturales Protegidasfacilitated legal and logistics aspects. Bell Labsgenerously donated the bait. We thank field crewsFrancisco Aguirre, Stacey Buckelew, MiguelDurazo, Luciana Luna, Luz María Salas-Flores,Arturo Ramírez, Antonio Ortiz, Anny Peralta,Tadeo Pfister, Noe Silva, Enrique Soqui, Juan A.Soriano, Kirsty Swinnerton, and Flor Torres forkeeping the enthusiasm under difficult fieldconditions. The GIS team integrated by AlejandroGonzález, Federico Méndez, and Montserrat Trigomade sure it was a very precise broadcast. ChristianLavoie helped with GIS protocols. Our partnersfrom Island Conservation Canada contributed to thedevelopment and execution of the project. KevinMiskel from Aspen Helicopters ran an excellentaerial operation. Helicopters Otago supplied the baitbucket and kindly shared their expertise. PrescottCol lege in Bahía Kino provided valuableinformation as well as logistics and field support.The Laboratorio de Pronóstico Meteorológico,Departamento de Oceanografía Física, CICESE,provided climate and forecast data. DOC andLandcare from New Zealand gave us advice andsuggestions. Funding was provided by PackardFoundation, Marisla Foundation, Fondo Mexicanopara la Conservacion de la Naturaleza, and other


foundat ions through Is land Conservat ion.Invaluable in-kind support was provided by theMexican Navy. Permits were granted by theWildlife General Direction (DGVS) of theE n v i r o n m e n t M i n i s t r y ( S E M A R N A T )(DGGIMAR.710/005985 and SGPA/DGVS/05248/07), SEGOB (DICOPPU/211/2331/07 andDICOPPU /2 11 / 24 24 /0 7 ) , COFEPRIS(06330041760001), and SCT (DGAC-PE-03-DTG3-FEI08). This paper improved thanks to twoanonymous reviewers. Christine Damiani kindlyimproved the English version.

REFERENCES

Aguirre-Muñoz, A., D. Croll, J. Donlan, R.W.Henry, M.A. Hermosillo, G. Howald, B. Keitt,L. Luna-Mendoza, M. Rodríguez-Malagón, L.M. Salas-Flores, A. Samaniego-Herrera, J.A.Sánchez-Pacheco, J. Sheppard, B. Tershy, J.Toro-Benito, S. Wolf, and B. Wood. 2008.High-impact conservation action: a case studyfrom the islands of western Mexico. Ambio37(2):101–107.

Atkinson, I.A.E. 1985. The spread of commensalspecies of Rattus to oceanic islands and theireffects on island avifaunas. Pages 35–81. In:Moors, P.J. (ed.), Conservation of Island Birds.International Council of Bird PreservationTech. Pub. 3.

Blackburn, T., P. Casey, R.P. Duncan, K. Evans,and K.J. Gaston. 2004. Avian extinction andmammalian introductions on oceanic islands.Science 35:1955–1958.

Campbell, E.W. 1991. The effect of introducedroof rats on bird diversity of Antillean Cays.Journal of Field Ornithology 62:343–348.

Case, T.J., M.L. Cody, and E. Ezcurra (eds.). 2002.A New Island Biogeography of the Sea ofCortéz. Oxford University Press, New York,669 pp.

CONAP. 2000. Programa de Manejo. Área deProtección de Flora y Fauna Islas del Golfo deCalifornia. México. SEMARNAT-CONANP.México, D.F., 262 pp.

CONANP. 2007. Programa de Conservación yManejo. Reserva de la Biósfera Isla San PedroMártir. México. SEMARNAT-CONANP.

Courchamp, F., J.L. Chapuis, and M. Pascal. 2003.Mammal invaders on islands: impact, controland control impact. Biological Reviews78:347–383.

Cree, A., C.H. Daugherty, and J.M. Hay. 1992.Reproduction of a rare New Zealand reptile,the tuatara Spenodon punctatus on rat-free andrat-inhabited islands. Conservation Biology2:373–383.

Frick, W.F., and P.A. Heady III. 2007. Monitoringthe Response of the Endemic Fish-eating Bat(Myotis vivesi) to Rat Eradication Efforts onIsla San Pedro Mártir, Sonora, México. Reportto Conservación de Islas. Central Coast BatResearch Group, CA.

González Bernal, M.A., E. Mellink, and C. PayánEzquerra. 2001a. Cnemidophorus tigris rangeextension. Herpetological Review 32(3):193.

González Bernal, M.A., E. Mellink, and C. PayánEzquerra. 2001b. Urosaurus ornatus rangeextension. Herpetological Review 32(3):193–194.

González Bernal, M.A., E. Mellink, and J.R. FongMendoza. 2002. Nesting birds of Farallón deSan Ignacio, Sinaloa. México. Western Birds33:254–257.

Grismer, L.L. 2002. Amphibians and reptiles ofBaja California, including its Pacific islandsand the islands in the sea of Cortes. Universityof California Press. Los Angeles.

Groombridge, B. 1992. World ConservationMonitoring Centre, British Museum (NaturalHistory), International Union for Conservationof Nature and Natural Resources. GlobalBiodiversity: Status of the Earth’s LivingResources: A Report. Chapman and Hall,London, UK.

Hoare, J.M., and K.M. Hare. 2006. The impact ofbrodifacoum on non-target wildlife: gaps inknowledge. New Zealand Journal of Ecology30(2):157–167.

Howald, G.R., K.R. Faulkner, B. Tershy, B. Keitt,H. Gellerman, E.M. Creel, M. Grinnel, S.Ortega, and D.A. Croll. 2005. Eradication ofblack rat from Anacapa Island: biological andsocial considerations. Pages 299–312. In:Garcelon, D.K., and C.A. Schwemm (eds.),Proceedings of the Sixth California IslandsSymposium. Institute for Wildlife Studies,Arcata, CA.


Howald, G., J. Donlan, J.P. Galván, J. Russell, J.Parkes, A. Samaniego, Y. Wang, D. Veitch, P.Genovesi, M. Pascal, A. Saunders, and B.Tershy. 2007. Invasive rodent eradication onislands. Conservation Biology 21(4):1021–1031.

Lee , K.G. , and J .C. Yoo. 2002. Breed ingpopu la t i on o f s t r e ake d s hea rw a te r s(Calonectris leucomelas) and the effect ofNorway rat (Rattus norvegicus) predation onSasudo Island. Journal of the YamashinaInstitute for Ornithology 33(2):142-147.

Mack R., D. Simberloff, W.M. Consdale, H.Evans, M. Clout, and F.A. Bazzaz. 2000.Biotic invasions: causes, epidemiology, globalconsequences and cont ro l . Ecologica lApplications 10:689–710.

Navarrete, S.A., and J.C. Castilla. 1993. Predationby Norway rats in the intertidal zone of centralChile. Marine Ecology Progress Series92:187–199.

Orueta, J., Y. Aranda, T. Gómez, G. Tapia and L.Sánchez-Marmol. 2005. Successful eradicationof invasive rodents from a small island throughpulsed bait ing inside covered stat ions.Biological Invasions 7:141–147.

Peralta-García, A., A. Samaniego-Herrera, and J.H.Valdez-Villavicencio. 2007. Registros nuevosde reptiles en islas del Noroeste de México.Acta Zoológica Mexicana (n.s.). 23(1):179–182.

Reaser, J.K., L.A. Meyerson, Q. Cronk, M.Poorter, L.G. Eldrege, E. Green, M. Kairo, P.Latasi, R.N. Mack, J. Mauremootoo, D.O’Dowd, W. Orapa, S. Sastroutomo, A.Saunders, C. Shine, S. Thrainsson, and L.Vaiutu. 2007. Ecological and socioeconomicimpacts of invasive alien species in islandecosystems. Environmental Conservation34(2):98–111.

Rodríguez, C., R. Torres, and H. Drummond. 2006.Eradicating introduced mammals from a

f o r e s t e d t r o p i c a l i s l a n d . B i o l o g i c a lConservation 130:98–105.

Tershy, B.R., and D. Breese. 1997. The birds ofSan Pedro Mártir island, Gulf of California.México. Western Birds 28:96–107.

Tershy, B.R., D. Breese, and D.A. Croll. 1997.Human perturbations and conservationstrategies for San Pedro Mártir Island, Gulf ofCa l i fo rn i a , Mé x ico . En v i ro nme n ta lConservation 24:261–270.

Tershy, B.R., C.J. Donlan, B.S. Keiit, D.A. Croll,J.A. Sanchez, B. Wood, M.A. Hermosillo, G.Howald, and N. Biavaschi. 2002. Islandconserva t ion in nor th-wes t Mexico : aconservation model integrating research,education, and exotic mammal eradication. In:Veitch, C.R., and M.N. Clout (eds.), Turningthe Tide: The Eradication of Invasive Species.IUCN. Switzerland.

Towns, D.R., I.A.E Atkinson, and C.H. Daugherty.2006. Have the harmful effects of introducedrats on island been exaggerated? BiologicalInvasions 8(4):863–891.

Towns, D.R., and K.G. Broome. 2003. From smallMaria to massive Campbell: forty years of rateradications from New Zealand islands. NewZealand Journal of Zoology 30:377–398.

Veitch, C.R., and M.N. Clout. 2002. Turning theTide: The Eradication of Invasive Species.IUCN. Switzerland.

V e l a r d e , E . , a n d D . W . A nd e r s o n . 1 9 9 4 .Conservation and management of seabirdislands in the Gulf of California: setbacks andsuccesses. In: Nettleship, D.N., J. Burger, andM. Gochfeld (eds), Seabirds on Islands: Treats,Case Studies and Action Plans. Cambridge,UK: ICBP Tech. Pub.

Whitaker, A.H. 1973. Lizard populations on islandwith and without Polynesian rats Rattusexulans. Proceedings of the New ZealandEcological Society 20:121–130.

Documents

ERADICATION OF BLACK RATS FROM FARALLÓN …iws.org/CISProceedings/7th_CIS_Proceedings/Samaniego-Herrera_et_al.pdferadication of black rats from farallÓn de san ignacio and san pedro